Yesterday
-
model card: plainer wording, credit upstream project only
2 days ago
-
final model card: 4B-token run, held-out story 1.312 / web 2.866
-
resume state step 122071
-
mid-run checkpoint step 122071 (4.00B tokens)
-
mid-run checkpoint step 115000 (3.77B tokens)
-
mid-run checkpoint step 105500 (3.46B tokens)
-
resume state step 96000
-
mid-run checkpoint step 96000 (3.15B tokens)
3 days ago
-
resume state step 94775
-
mid-run checkpoint step 94775 (3.11B tokens)
-
mid-run checkpoint step 86000 (2.82B tokens)
-
mid-run checkpoint step 76500 (2.51B tokens)
-
resume state step 66500
-
mid-run checkpoint step 66500 (2.18B tokens)
-
mid-run checkpoint step 57500 (1.88B tokens)
-
mid-run checkpoint step 48000 (1.57B tokens)
4 days ago
-
resume state: step 46013, 1.51B tokens
-
mid-run checkpoint step 46013 (1.51B tokens)
-
mid-run checkpoint step 45500 (1.49B tokens)
-
mid-run checkpoint step 36000 (1.18B tokens)
-
mid-run checkpoint step 26500 (0.87B tokens)
-
mid-run checkpoint step 17000 (0.56B tokens)
-
mid-run checkpoint step 8000 (0.26B tokens)
-
initial commit
